Assessing the Performance of Deep Learning for Automated Gleason Grading in Prostate Cancer.

Prostate cancer is a dominant health concern calling for advanced diagnostic tools. Utilizing digital pathology and artificial intelligence, this study explores the potential of 11 deep neural network architectures for automated Gleason grading in prostate carcinoma focusing on comparing traditional and recent architectures. A standardized image classification pipeline, based on the AUCMEDI framework, facilitated robust evaluation using an in-house dataset consisting of 34,264 annotated tissue tiles. The results indicated varying sensitivity across architectures, with ConvNeXt demonstrating the strongest performance. Notably, newer architectures achieved superior performance, even though with challenges in differentiating closely related Gleason grades. The ConvNeXt model was capable of learning a balance between complexity and generalizability. Overall, this study lays the groundwork for enhanced Gleason grading systems, potentially improving diagnostic efficiency for prostate cancer.

Endoplasmic reticulum-bound ANAC013 factor is cleaved by RHOMBOID-LIKE 2 during the initial response to hypoxia in Arabidopsis thaliana.

Aerobic reactions are essential to sustain plant growth and development. Impaired oxygen availability due to excessive water availability, e.g., during waterlogging or flooding, reduces plant productivity and survival. Consequently, plants monitor oxygen availability to adjust growth and metabolism accordingly. Despite the identification of central components in hypoxia adaptation in recent years, molecular pathways involved in the very early activation of low-oxygen responses are insufficiently understood. Here, we characterized three endoplasmic reticulum (ER)-anchored Arabidopsis ANAC transcription factors, namely ANAC013, ANAC016, and ANAC017, which bind to the promoters of a subset of hypoxia core genes (HCGs) and activate their expression. However, only ANAC013 translocates to the nucleus at the onset of hypoxia, i.e., after 1.5 h of stress. Upon hypoxia, nuclear ANAC013 associates with the promoters of multiple HCGs. Mechanistically, we identified residues in the transmembrane domain of ANAC013 to be essential for transcription factor release from the ER, and provide evidence that RHOMBOID-LIKE 2 (RBL2) protease mediates ANAC013 release under hypoxia. Release of ANAC013 by RBL2 also occurs upon mitochondrial dysfunction. Consistently, like ANAC013 knockdown lines, rbl knockout mutants exhibit impaired low-oxygen tolerance. Taken together, we uncovered an ER-localized ANAC013-RBL2 module, which is active during the initial phase of hypoxia to enable fast transcriptional reprogramming.

Proteolytic Activation of Plant Membrane-Bound Transcription Factors.

Due to the presence of a transmembrane domain, the subcellular mobility plan of membrane-bound or membrane-tethered transcription factors (MB-TFs) differs from that of their cytosolic counterparts. The MB-TFs are mostly locked in (sub)cellular membranes, until they are released by a proteolytic cleavage event or when the transmembrane domain (TMD) is omitted from the transcript due to alternative splicing. Here, we review the current knowledge on the proteolytic activation mechanisms of MB-TFs in plants, with a particular focus on regulated intramembrane proteolysis (RIP), and discuss the analogy with the proteolytic cleavage of MB-TFs in animal systems. We present a comprehensive inventory of all known and predicted MB-TFs in the model plant Arabidopsis thaliana and examine their experimentally determined or anticipated subcellular localizations and membrane topologies. We predict proteolytically activated MB-TFs by the mapping of protease recognition sequences and structural features that facilitate RIP in and around the TMD, based on data from metazoan intramembrane proteases. Finally, the MB-TF functions in plant responses to environmental stresses and in plant development are considered and novel functions for still uncharacterized MB-TFs are forecasted by means of a regulatory network-based approach.